Index: llvm/include/llvm/CodeGen/ValueTypes.td
===================================================================
--- llvm/include/llvm/CodeGen/ValueTypes.td
+++ llvm/include/llvm/CodeGen/ValueTypes.td
@@ -26,8 +26,8 @@
 def i64     : ValueType<64,  6>;  // 64-bit integer value
 def i128    : ValueType<128, 7>;  // 128-bit integer value
 
-def bf16    : ValueType<16,   8>;  // 16-bit brain floating point value
-def f16     : ValueType<16,   9>;  // 16-bit floating point value
+def bf16    : ValueType<16,  8>;  // 16-bit brain floating point value
+def f16     : ValueType<16,  9>;  // 16-bit floating point value
 def f32     : ValueType<32,  10>;  // 32-bit floating point value
 def f64     : ValueType<64,  11>;  // 64-bit floating point value
 def f80     : ValueType<80,  12>;  // 80-bit floating point value
@@ -46,178 +46,184 @@
 def v512i1  : ValueType<512,  24>;  //  512 x i1 vector value
 def v1024i1 : ValueType<1024, 25>;  // 1024 x i1 vector value
 
-def v1i8    : ValueType<8,    26>;  //    1 x i8 vector value
-def v2i8    : ValueType<16,   27>;  //    2 x i8 vector value
-def v4i8    : ValueType<32,   28>;  //    4 x i8 vector value
-def v8i8    : ValueType<64,   29>;  //    8 x i8 vector value
-def v16i8   : ValueType<128,  30>;  //   16 x i8 vector value
-def v32i8   : ValueType<256,  31>;  //   32 x i8 vector value
-def v64i8   : ValueType<512,  32>;  //   64 x i8 vector value
-def v128i8  : ValueType<1024, 33>;  //  128 x i8 vector value
-def v256i8  : ValueType<2048, 34>;  //  256 x i8 vector value
-def v512i8  : ValueType<4096, 35>;  //  512 x i8 vector value
-def v1024i8 : ValueType<8192, 36>;  // 1024 x i8 vector value
-
-def v1i16   : ValueType<16,   37>;  //   1 x i16 vector value
-def v2i16   : ValueType<32,   38>;  //   2 x i16 vector value
-def v3i16   : ValueType<48,   39>;  //   3 x i16 vector value
-def v4i16   : ValueType<64,   40>;  //   4 x i16 vector value
-def v8i16   : ValueType<128,  41>;  //   8 x i16 vector value
-def v16i16  : ValueType<256,  42>;  //  16 x i16 vector value
-def v32i16  : ValueType<512,  43>;  //  32 x i16 vector value
-def v64i16  : ValueType<1024, 44>;  //  64 x i16 vector value
-def v128i16 : ValueType<2048, 45>;  // 128 x i16 vector value
-def v256i16 : ValueType<4096, 46>;  // 256 x i16 vector value
-def v512i16 : ValueType<8192, 47>;  // 512 x i16 vector value
-
-def v1i32    : ValueType<32,    48>;  //    1 x i32 vector value
-def v2i32    : ValueType<64,    49>;  //    2 x i32 vector value
-def v3i32    : ValueType<96,    50>;  //    3 x i32 vector value
-def v4i32    : ValueType<128,   51>;  //    4 x i32 vector value
-def v5i32    : ValueType<160,   52>;  //    5 x i32 vector value
-def v6i32    : ValueType<192,   53>;  //    6 x f32 vector value
-def v7i32    : ValueType<224,   54>;  //    7 x f32 vector value
-def v8i32    : ValueType<256,   55>;  //    8 x i32 vector value
-def v16i32   : ValueType<512,   56>;  //   16 x i32 vector value
-def v32i32   : ValueType<1024,  57>;  //   32 x i32 vector value
-def v64i32   : ValueType<2048,  58>;  //   64 x i32 vector value
-def v128i32  : ValueType<4096,  59>;  //  128 x i32 vector value
-def v256i32  : ValueType<8192,  60>;  //  256 x i32 vector value
-def v512i32  : ValueType<16384, 61>;  //  512 x i32 vector value
-def v1024i32 : ValueType<32768, 62>;  // 1024 x i32 vector value
-def v2048i32 : ValueType<65536, 63>;  // 2048 x i32 vector value
-
-def v1i64   : ValueType<64,    64>;  //   1 x i64 vector value
-def v2i64   : ValueType<128,   65>;  //   2 x i64 vector value
-def v3i64   : ValueType<192,   66>;  //   3 x i64 vector value
-def v4i64   : ValueType<256,   67>;  //   4 x i64 vector value
-def v8i64   : ValueType<512,   68>;  //   8 x i64 vector value
-def v16i64  : ValueType<1024,  69>;  //  16 x i64 vector value
-def v32i64  : ValueType<2048,  70>;  //  32 x i64 vector value
-def v64i64  : ValueType<4096,  71>;  //  64 x i64 vector value
-def v128i64 : ValueType<8192,  72>;  // 128 x i64 vector value
-def v256i64 : ValueType<16384, 73>;  // 256 x i64 vector value
-
-def v1i128 : ValueType<128, 74>;  //  1 x i128 vector value
-
-def v1f16    : ValueType<16,     75>;  //    1 x f16 vector value
-def v2f16    : ValueType<32,     76>;  //    2 x f16 vector value
-def v3f16    : ValueType<48,     77>;  //    3 x f16 vector value
-def v4f16    : ValueType<64,     78>;  //    4 x f16 vector value
-def v8f16    : ValueType<128,    79>;  //    8 x f16 vector value
-def v16f16   : ValueType<256,    80>;  //   16 x f16 vector value
-def v32f16   : ValueType<512,    81>;  //   32 x f16 vector value
-def v64f16   : ValueType<1024,   82>;  //   64 x f16 vector value
-def v128f16  : ValueType<2048,   83>;  //  128 x f16 vector value
-def v256f16  : ValueType<4096,   84>;  //  256 x f16 vector value
-def v512f16  : ValueType<8192,   85>;  //  512 x f16 vector value
-
-def v2bf16   : ValueType<32,     86>;  //    2 x bf16 vector value
-def v3bf16   : ValueType<48,     87>;  //    3 x bf16 vector value
-def v4bf16   : ValueType<64,     88>;  //    4 x bf16 vector value
-def v8bf16   : ValueType<128,    89>;  //    8 x bf16 vector value
-def v16bf16  : ValueType<256,    90>;  //   16 x bf16 vector value
-def v32bf16  : ValueType<512,    91>;  //   32 x bf16 vector value
-def v64bf16  : ValueType<1024,   92>;  //   64 x bf16 vector value
-def v128bf16 : ValueType<2048,   93>;  //  128 x bf16 vector value
-
-def v1f32    : ValueType<32,     94>;  //    1 x f32 vector value
-def v2f32    : ValueType<64,     95>;  //    2 x f32 vector value
-def v3f32    : ValueType<96,     96>;  //    3 x f32 vector value
-def v4f32    : ValueType<128,    97>;  //    4 x f32 vector value
-def v5f32    : ValueType<160,    98>;  //    5 x f32 vector value
-def v6f32    : ValueType<192,    99>;  //    6 x f32 vector value
-def v7f32    : ValueType<224,   100>;  //    7 x f32 vector value
-def v8f32    : ValueType<256,   101>;  //    8 x f32 vector value
-def v16f32   : ValueType<512,   102>;  //   16 x f32 vector value
-def v32f32   : ValueType<1024,  103>;  //   32 x f32 vector value
-def v64f32   : ValueType<2048,  104>;  //   64 x f32 vector value
-def v128f32  : ValueType<4096,  105>;  //  128 x f32 vector value
-def v256f32  : ValueType<8192,  106>;  //  256 x f32 vector value
-def v512f32  : ValueType<16384, 107>;  //  512 x f32 vector value
-def v1024f32 : ValueType<32768, 108>;  // 1024 x f32 vector value
-def v2048f32 : ValueType<65536, 109>;  // 2048 x f32 vector value
-
-def v1f64    : ValueType<64,    110>;  //    1 x f64 vector value
-def v2f64    : ValueType<128,   111>;  //    2 x f64 vector value
-def v3f64    : ValueType<192,   112>;  //    3 x f64 vector value
-def v4f64    : ValueType<256,   113>;  //    4 x f64 vector value
-def v8f64    : ValueType<512,   114>;  //    8 x f64 vector value
-def v16f64   : ValueType<1024,  115>;  //   16 x f64 vector value
-def v32f64   : ValueType<2048,  116>;  //   32 x f64 vector value
-def v64f64   : ValueType<4096,  117>;  //   64 x f64 vector value
-def v128f64  : ValueType<8192,  118>;  //  128 x f64 vector value
-def v256f64  : ValueType<16384, 119>;  //  256 x f64 vector value
-
-def nxv1i1  : ValueType<1,  120>;  // n x  1 x i1  vector value
-def nxv2i1  : ValueType<2,  121>;  // n x  2 x i1  vector value
-def nxv4i1  : ValueType<4,  122>;  // n x  4 x i1  vector value
-def nxv8i1  : ValueType<8,  123>;  // n x  8 x i1  vector value
-def nxv16i1 : ValueType<16, 124>;  // n x 16 x i1  vector value
-def nxv32i1 : ValueType<32, 125>;  // n x 32 x i1  vector value
-def nxv64i1 : ValueType<64, 126>;  // n x 64 x i1  vector value
-
-def nxv1i8  : ValueType<8,   127>;  // n x  1 x i8  vector value
-def nxv2i8  : ValueType<16,  128>;  // n x  2 x i8  vector value
-def nxv4i8  : ValueType<32,  129>;  // n x  4 x i8  vector value
-def nxv8i8  : ValueType<64,  130>;  // n x  8 x i8  vector value
-def nxv16i8 : ValueType<128, 131>;  // n x 16 x i8  vector value
-def nxv32i8 : ValueType<256, 132>;  // n x 32 x i8  vector value
-def nxv64i8 : ValueType<512, 133>;  // n x 64 x i8  vector value
-
-def nxv1i16  : ValueType<16,  134>;  // n x  1 x i16 vector value
-def nxv2i16  : ValueType<32,  135>;  // n x  2 x i16 vector value
-def nxv4i16  : ValueType<64,  136>;  // n x  4 x i16 vector value
-def nxv8i16  : ValueType<128, 137>;  // n x  8 x i16 vector value
-def nxv16i16 : ValueType<256, 138>;  // n x 16 x i16 vector value
-def nxv32i16 : ValueType<512, 139>;  // n x 32 x i16 vector value
-
-def nxv1i32  : ValueType<32,   140>;  // n x  1 x i32 vector value
-def nxv2i32  : ValueType<64,   141>;  // n x  2 x i32 vector value
-def nxv4i32  : ValueType<128,  142>;  // n x  4 x i32 vector value
-def nxv8i32  : ValueType<256,  143>;  // n x  8 x i32 vector value
-def nxv16i32 : ValueType<512,  144>;  // n x 16 x i32 vector value
-def nxv32i32 : ValueType<1024, 145>;  // n x 32 x i32 vector value
-
-def nxv1i64  : ValueType<64,   146>;  // n x  1 x i64 vector value
-def nxv2i64  : ValueType<128,  147>;  // n x  2 x i64 vector value
-def nxv4i64  : ValueType<256,  148>;  // n x  4 x i64 vector value
-def nxv8i64  : ValueType<512,  149>;  // n x  8 x i64 vector value
-def nxv16i64 : ValueType<1024, 150>;  // n x 16 x i64 vector value
-def nxv32i64 : ValueType<2048, 151>;  // n x 32 x i64 vector value
-
-def nxv1f16  : ValueType<16,  152>;  // n x  1 x  f16 vector value
-def nxv2f16  : ValueType<32,  153>;  // n x  2 x  f16 vector value
-def nxv4f16  : ValueType<64,  154>;  // n x  4 x  f16 vector value
-def nxv8f16  : ValueType<128, 155>;  // n x  8 x  f16 vector value
-def nxv16f16 : ValueType<256, 156>;  // n x 16 x  f16 vector value
-def nxv32f16 : ValueType<512, 157>;  // n x 32 x  f16 vector value
-
-def nxv1bf16 : ValueType<16,  158>;  // n x  1 x bf16 vector value
-def nxv2bf16 : ValueType<32,  159>;  // n x  2 x bf16 vector value
-def nxv4bf16 : ValueType<64,  160>;  // n x  4 x bf16 vector value
-def nxv8bf16 : ValueType<128, 161>;  // n x  8 x bf16 vector value
-
-def nxv1f32  : ValueType<32,  162>;  // n x  1 x  f32 vector value
-def nxv2f32  : ValueType<64,  163>;  // n x  2 x  f32 vector value
-def nxv4f32  : ValueType<128, 164>;  // n x  4 x  f32 vector value
-def nxv8f32  : ValueType<256, 165>;  // n x  8 x  f32 vector value
-def nxv16f32 : ValueType<512, 166>;  // n x 16 x  f32 vector value
-
-def nxv1f64  : ValueType<64,  167>;  // n x  1 x  f64 vector value
-def nxv2f64  : ValueType<128, 168>;  // n x  2 x  f64 vector value
-def nxv4f64  : ValueType<256, 169>;  // n x  4 x  f64 vector value
-def nxv8f64  : ValueType<512, 170>;  // n x  8 x  f64 vector value
-
-def x86mmx    : ValueType<64,   171>;  // X86 MMX value
-def FlagVT    : ValueType<0,    172>;  // Pre-RA sched glue
-def isVoid    : ValueType<0,    173>;  // Produces no value
-def untyped   : ValueType<8,    174>;  // Produces an untyped value
-def funcref   : ValueType<0,    175>;  // WebAssembly's funcref type
-def externref : ValueType<0,    176>;  // WebAssembly's externref type
-def x86amx    : ValueType<8192, 177>;  // X86 AMX value
-def i64x8     : ValueType<512,  178>;  // 8 Consecutive GPRs (AArch64)
-
+def v128i2  : ValueType<256,  26>;   //  128 x i2 vector value
+
+def v64i4   : ValueType<256,  27>;   //   64 x i4 vector value
+
+def v1i8    : ValueType<8,    28>;  //    1 x i8 vector value
+def v2i8    : ValueType<16,   29>;  //    2 x i8 vector value
+def v4i8    : ValueType<32,   30>;  //    4 x i8 vector value
+def v8i8    : ValueType<64,   31>;  //    8 x i8 vector value
+def v16i8   : ValueType<128,  32>;  //   16 x i8 vector value
+def v32i8   : ValueType<256,  33>;  //   32 x i8 vector value
+def v64i8   : ValueType<512,  34>;  //   64 x i8 vector value
+def v128i8  : ValueType<1024, 35>;  //  128 x i8 vector value
+def v256i8  : ValueType<2048, 36>;  //  256 x i8 vector value
+def v512i8  : ValueType<4096, 37>;  //  512 x i8 vector value
+def v1024i8 : ValueType<8192, 38>;  // 1024 x i8 vector value
+
+def v1i16   : ValueType<16,   39>;  //   1 x i16 vector value
+def v2i16   : ValueType<32,   40>;  //   2 x i16 vector value
+def v3i16   : ValueType<48,   41>;  //   3 x i16 vector value
+def v4i16   : ValueType<64,   42>;  //   4 x i16 vector value
+def v8i16   : ValueType<128,  43>;  //   8 x i16 vector value
+def v16i16  : ValueType<256,  44>;  //  16 x i16 vector value
+def v32i16  : ValueType<512,  45>;  //  32 x i16 vector value
+def v64i16  : ValueType<1024, 46>;  //  64 x i16 vector value
+def v128i16 : ValueType<2048, 47>;  // 128 x i16 vector value
+def v256i16 : ValueType<4096, 48>;  // 256 x i16 vector value
+def v512i16 : ValueType<8192, 49>;  // 512 x i16 vector value
+
+def v1i32    : ValueType<32,    50>;  //    1 x i32 vector value
+def v2i32    : ValueType<64,    51>;  //    2 x i32 vector value
+def v3i32    : ValueType<96,    52>;  //    3 x i32 vector value
+def v4i32    : ValueType<128,   53>;  //    4 x i32 vector value
+def v5i32    : ValueType<160,   54>;  //    5 x i32 vector value
+def v6i32    : ValueType<192,   55>;  //    6 x f32 vector value
+def v7i32    : ValueType<224,   56>;  //    7 x f32 vector value
+def v8i32    : ValueType<256,   57>;  //    8 x i32 vector value
+def v16i32   : ValueType<512,   58>;  //   16 x i32 vector value
+def v32i32   : ValueType<1024,  59>;  //   32 x i32 vector value
+def v64i32   : ValueType<2048,  60>;  //   64 x i32 vector value
+def v128i32  : ValueType<4096,  61>;  //  128 x i32 vector value
+def v256i32  : ValueType<8192,  62>;  //  256 x i32 vector value
+def v512i32  : ValueType<16384, 63>;  //  512 x i32 vector value
+def v1024i32 : ValueType<32768, 64>;  // 1024 x i32 vector value
+def v2048i32 : ValueType<65536, 65>;  // 2048 x i32 vector value
+
+def v1i64   : ValueType<64,    66>;  //   1 x i64 vector value
+def v2i64   : ValueType<128,   67>;  //   2 x i64 vector value
+def v3i64   : ValueType<192,   69>;  //   3 x i64 vector value
+def v4i64   : ValueType<256,   69>;  //   4 x i64 vector value
+def v8i64   : ValueType<512,   70>;  //   8 x i64 vector value
+def v16i64  : ValueType<1024,  71>;  //  16 x i64 vector value
+def v32i64  : ValueType<2048,  71>;  //  32 x i64 vector value
+def v64i64  : ValueType<4096,  72>;  //  64 x i64 vector value
+def v128i64 : ValueType<8192,  73>;  // 128 x i64 vector value
+def v256i64 : ValueType<16384, 74>;  // 256 x i64 vector value
+
+def v1i128 : ValueType<128, 76>;  //  1 x i128 vector value
+
+def v1f16    : ValueType<16,     77>;  //    1 x f16 vector value
+def v2f16    : ValueType<32,     78>;  //    2 x f16 vector value
+def v3f16    : ValueType<48,     79>;  //    3 x f16 vector value
+def v4f16    : ValueType<64,     80>;  //    4 x f16 vector value
+def v8f16    : ValueType<128,    81>;  //    8 x f16 vector value
+def v16f16   : ValueType<256,    82>;  //   16 x f16 vector value
+def v32f16   : ValueType<512,    83>;  //   32 x f16 vector value
+def v64f16   : ValueType<1024,   84>;  //   64 x f16 vector value
+def v128f16  : ValueType<2048,   85>;  //  128 x f16 vector value
+def v256f16  : ValueType<4096,   86>;  //  256 x f16 vector value
+def v512f16  : ValueType<8192,   87>;  //  512 x f16 vector value
+
+def v2bf16   : ValueType<32,     88>;  //    2 x bf16 vector value
+def v3bf16   : ValueType<48,     89>;  //    3 x bf16 vector value
+def v4bf16   : ValueType<64,     90>;  //    4 x bf16 vector value
+def v8bf16   : ValueType<128,    91>;  //    8 x bf16 vector value
+def v16bf16  : ValueType<256,    92>;  //   16 x bf16 vector value
+def v32bf16  : ValueType<512,    93>;  //   32 x bf16 vector value
+def v64bf16  : ValueType<1024,   94>;  //   64 x bf16 vector value
+def v128bf16 : ValueType<2048,   95>;  //  128 x bf16 vector value
+
+def v1f32    : ValueType<32,     96>;  //    1 x f32 vector value
+def v2f32    : ValueType<64,     97>;  //    2 x f32 vector value
+def v3f32    : ValueType<96,     98>;  //    3 x f32 vector value
+def v4f32    : ValueType<128,    99>;  //    4 x f32 vector value
+def v5f32    : ValueType<160,   100>;  //    5 x f32 vector value
+def v6f32    : ValueType<192,   101>;  //    6 x f32 vector value
+def v7f32    : ValueType<224,   102>;  //    7 x f32 vector value
+def v8f32    : ValueType<256,   103>;  //    8 x f32 vector value
+def v16f32   : ValueType<512,   104>;  //   16 x f32 vector value
+def v32f32   : ValueType<1024,  105>;  //   32 x f32 vector value
+def v64f32   : ValueType<2048,  106>;  //   64 x f32 vector value
+def v128f32  : ValueType<4096,  107>;  //  128 x f32 vector value
+def v256f32  : ValueType<8192,  108>;  //  256 x f32 vector value
+def v512f32  : ValueType<16384, 109>;  //  512 x f32 vector value
+def v1024f32 : ValueType<32768, 110>;  // 1024 x f32 vector value
+def v2048f32 : ValueType<65536, 111>;  // 2048 x f32 vector value
+
+def v1f64    : ValueType<64,    112>;  //    1 x f64 vector value
+def v2f64    : ValueType<128,   113>;  //    2 x f64 vector value
+def v3f64    : ValueType<192,   114>;  //    3 x f64 vector value
+def v4f64    : ValueType<256,   115>;  //    4 x f64 vector value
+def v8f64    : ValueType<512,   116>;  //    8 x f64 vector value
+def v16f64   : ValueType<1024,  117>;  //   16 x f64 vector value
+def v32f64   : ValueType<2048,  118>;  //   32 x f64 vector value
+def v64f64   : ValueType<4096,  119>;  //   64 x f64 vector value
+def v128f64  : ValueType<8192,  120>;  //  128 x f64 vector value
+def v256f64  : ValueType<16384, 121>;  //  256 x f64 vector value
+
+def nxv1i1  : ValueType<1,  122>;  // n x  1 x i1  vector value
+def nxv2i1  : ValueType<2,  123>;  // n x  2 x i1  vector value
+def nxv4i1  : ValueType<4,  124>;  // n x  4 x i1  vector value
+def nxv8i1  : ValueType<8,  125>;  // n x  8 x i1  vector value
+def nxv16i1 : ValueType<16, 126>;  // n x 16 x i1  vector value
+def nxv32i1 : ValueType<32, 127>;  // n x 32 x i1  vector value
+def nxv64i1 : ValueType<64, 128>;  // n x 64 x i1  vector value
+
+def nxv1i8  : ValueType<8,   129>;  // n x  1 x i8  vector value
+def nxv2i8  : ValueType<16,  130>;  // n x  2 x i8  vector value
+def nxv4i8  : ValueType<32,  131>;  // n x  4 x i8  vector value
+def nxv8i8  : ValueType<64,  132>;  // n x  8 x i8  vector value
+def nxv16i8 : ValueType<128, 133>;  // n x 16 x i8  vector value
+def nxv32i8 : ValueType<256, 134>;  // n x 32 x i8  vector value
+def nxv64i8 : ValueType<512, 135>;  // n x 64 x i8  vector value
+
+def nxv1i16  : ValueType<16,  136>;  // n x  1 x i16 vector value
+def nxv2i16  : ValueType<32,  137>;  // n x  2 x i16 vector value
+def nxv4i16  : ValueType<64,  138>;  // n x  4 x i16 vector value
+def nxv8i16  : ValueType<128, 139>;  // n x  8 x i16 vector value
+def nxv16i16 : ValueType<256, 140>;  // n x 16 x i16 vector value
+def nxv32i16 : ValueType<512, 141>;  // n x 32 x i16 vector value
+
+def nxv1i32  : ValueType<32,   142>;  // n x  1 x i32 vector value
+def nxv2i32  : ValueType<64,   143>;  // n x  2 x i32 vector value
+def nxv4i32  : ValueType<128,  144>;  // n x  4 x i32 vector value
+def nxv8i32  : ValueType<256,  145>;  // n x  8 x i32 vector value
+def nxv16i32 : ValueType<512,  146>;  // n x 16 x i32 vector value
+def nxv32i32 : ValueType<1024, 147>;  // n x 32 x i32 vector value
+
+def nxv1i64  : ValueType<64,   148>;  // n x  1 x i64 vector value
+def nxv2i64  : ValueType<128,  149>;  // n x  2 x i64 vector value
+def nxv4i64  : ValueType<256,  150>;  // n x  4 x i64 vector value
+def nxv8i64  : ValueType<512,  151>;  // n x  8 x i64 vector value
+def nxv16i64 : ValueType<1024, 152>;  // n x 16 x i64 vector value
+def nxv32i64 : ValueType<2048, 153>;  // n x 32 x i64 vector value
+
+def nxv1f16  : ValueType<16,  154>;  // n x  1 x  f16 vector value
+def nxv2f16  : ValueType<32,  155>;  // n x  2 x  f16 vector value
+def nxv4f16  : ValueType<64,  156>;  // n x  4 x  f16 vector value
+def nxv8f16  : ValueType<128, 157>;  // n x  8 x  f16 vector value
+def nxv16f16 : ValueType<256, 158>;  // n x 16 x  f16 vector value
+def nxv32f16 : ValueType<512, 159>;  // n x 32 x  f16 vector value
+
+def nxv1bf16 : ValueType<16,  160>;  // n x  1 x bf16 vector value
+def nxv2bf16 : ValueType<32,  161>;  // n x  2 x bf16 vector value
+def nxv4bf16 : ValueType<64,  162>;  // n x  4 x bf16 vector value
+def nxv8bf16 : ValueType<128, 163>;  // n x  8 x bf16 vector value
+
+def nxv1f32  : ValueType<32,  164>;  // n x  1 x  f32 vector value
+def nxv2f32  : ValueType<64,  165>;  // n x  2 x  f32 vector value
+def nxv4f32  : ValueType<128, 166>;  // n x  4 x  f32 vector value
+def nxv8f32  : ValueType<256, 167>;  // n x  8 x  f32 vector value
+def nxv16f32 : ValueType<512, 168>;  // n x 16 x  f32 vector value
+
+def nxv1f64  : ValueType<64,  169>;  // n x  1 x  f64 vector value
+def nxv2f64  : ValueType<128, 170>;  // n x  2 x  f64 vector value
+def nxv4f64  : ValueType<256, 171>;  // n x  4 x  f64 vector value
+def nxv8f64  : ValueType<512, 172>;  // n x  8 x  f64 vector value
+
+def x86mmx    : ValueType<64,   173>;  // X86 MMX value
+def FlagVT    : ValueType<0,    174>;  // Pre-RA sched glue
+def isVoid    : ValueType<0,    175>;  // Produces no value
+def untyped   : ValueType<8,    176>;  // Produces an untyped value
+def funcref   : ValueType<0,    177>;  // WebAssembly's funcref type
+def externref : ValueType<0,    178>;  // WebAssembly's externref type
+def x86amx    : ValueType<8192, 179>;  // X86 AMX value
+def i64x8     : ValueType<512,  180>;  // 8 Consecutive GPRs (AArch64)
+
+def i2        : ValueType<2,    181>;  // 2-bit boolean value
+def i4        : ValueType<4,    182>;  // 4-bit boolean value
 
 def token      : ValueType<0, 248>;  // TokenTy
 def MetadataVT : ValueType<0, 249>;  // Metadata
Index: llvm/include/llvm/Support/MachineValueType.h
===================================================================
--- llvm/include/llvm/Support/MachineValueType.h
+++ llvm/include/llvm/Support/MachineValueType.h
@@ -73,111 +73,115 @@
       v512i1         =  24,   //  512 x i1
       v1024i1        =  25,   // 1024 x i1
 
-      v1i8           =  26,   //    1 x i8
-      v2i8           =  27,   //    2 x i8
-      v4i8           =  28,   //    4 x i8
-      v8i8           =  29,   //    8 x i8
-      v16i8          =  30,   //   16 x i8
-      v32i8          =  31,   //   32 x i8
-      v64i8          =  32,   //   64 x i8
-      v128i8         =  33,   //  128 x i8
-      v256i8         =  34,   //  256 x i8
-      v512i8         =  35,   //  512 x i8
-      v1024i8        =  36,   // 1024 x i8
-
-      v1i16          =  37,   //   1 x i16
-      v2i16          =  38,   //   2 x i16
-      v3i16          =  39,   //   3 x i16
-      v4i16          =  40,   //   4 x i16
-      v8i16          =  41,   //   8 x i16
-      v16i16         =  42,   //  16 x i16
-      v32i16         =  43,   //  32 x i16
-      v64i16         =  44,   //  64 x i16
-      v128i16        =  45,   // 128 x i16
-      v256i16        =  46,   // 256 x i16
-      v512i16        =  47,   // 512 x i16
-
-      v1i32          =  48,   //    1 x i32
-      v2i32          =  49,   //    2 x i32
-      v3i32          =  50,   //    3 x i32
-      v4i32          =  51,   //    4 x i32
-      v5i32          =  52,   //    5 x i32
-      v6i32          =  53,   //    6 x i32
-      v7i32          =  54,   //    7 x i32
-      v8i32          =  55,   //    8 x i32
-      v16i32         =  56,   //   16 x i32
-      v32i32         =  57,   //   32 x i32
-      v64i32         =  58,   //   64 x i32
-      v128i32        =  59,   //  128 x i32
-      v256i32        =  60,   //  256 x i32
-      v512i32        =  61,   //  512 x i32
-      v1024i32       =  62,   // 1024 x i32
-      v2048i32       =  63,   // 2048 x i32
-
-      v1i64          =  64,   //   1 x i64
-      v2i64          =  65,   //   2 x i64
-      v3i64          =  66,   //   3 x i64
-      v4i64          =  67,   //   4 x i64
-      v8i64          =  68,   //   8 x i64
-      v16i64         =  69,   //  16 x i64
-      v32i64         =  70,   //  32 x i64
-      v64i64         =  71,   //  64 x i64
-      v128i64        =  72,   // 128 x i64
-      v256i64        =  73,   // 256 x i64
-
-      v1i128         =  74,   //  1 x i128
+      v128i2         =  26,   //  128 x i2
+
+      v64i4          =  27,   //   64 x i4
+
+      v1i8           =  28,   //    1 x i8
+      v2i8           =  29,   //    2 x i8
+      v4i8           =  30,   //    4 x i8
+      v8i8           =  31,   //    8 x i8
+      v16i8          =  32,   //   16 x i8
+      v32i8          =  33,   //   32 x i8
+      v64i8          =  34,   //   64 x i8
+      v128i8         =  35,   //  128 x i8
+      v256i8         =  36,   //  256 x i8
+      v512i8         =  37,   //  512 x i8
+      v1024i8        =  38,   // 1024 x i8
+
+      v1i16          =  39,   //   1 x i16
+      v2i16          =  40,   //   2 x i16
+      v3i16          =  41,   //   3 x i16
+      v4i16          =  42,   //   4 x i16
+      v8i16          =  43,   //   8 x i16
+      v16i16         =  44,   //  16 x i16
+      v32i16         =  45,   //  32 x i16
+      v64i16         =  46,   //  64 x i16
+      v128i16        =  47,   // 128 x i16
+      v256i16        =  48,   // 256 x i16
+      v512i16        =  49,   // 512 x i16
+
+      v1i32          =  50,   //    1 x i32
+      v2i32          =  51,   //    2 x i32
+      v3i32          =  52,   //    3 x i32
+      v4i32          =  53,   //    4 x i32
+      v5i32          =  54,   //    5 x i32
+      v6i32          =  55,   //    6 x i32
+      v7i32          =  56,   //    7 x i32
+      v8i32          =  57,   //    8 x i32
+      v16i32         =  58,   //   16 x i32
+      v32i32         =  59,   //   32 x i32
+      v64i32         =  60,   //   64 x i32
+      v128i32        =  61,   //  128 x i32
+      v256i32        =  62,   //  256 x i32
+      v512i32        =  63,   //  512 x i32
+      v1024i32       =  64,   // 1024 x i32
+      v2048i32       =  65,   // 2048 x i32
+
+      v1i64          =  66,   //   1 x i64
+      v2i64          =  67,   //   2 x i64
+      v3i64          =  68,   //   3 x i64
+      v4i64          =  69,   //   4 x i64
+      v8i64          =  70,   //   8 x i64
+      v16i64         =  71,   //  16 x i64
+      v32i64         =  72,   //  32 x i64
+      v64i64         =  73,   //  64 x i64
+      v128i64        =  74,   // 128 x i64
+      v256i64        =  75,   // 256 x i64
+
+      v1i128         =  76,   //  1 x i128
 
       FIRST_INTEGER_FIXEDLEN_VECTOR_VALUETYPE = v1i1,
       LAST_INTEGER_FIXEDLEN_VECTOR_VALUETYPE = v1i128,
 
-      v1f16          =  75,   //    1 x f16
-      v2f16          =  76,   //    2 x f16
-      v3f16          =  77,   //    3 x f16
-      v4f16          =  78,   //    4 x f16
-      v8f16          =  79,   //    8 x f16
-      v16f16         =  80,   //   16 x f16
-      v32f16         =  81,   //   32 x f16
-      v64f16         =  82,   //   64 x f16
-      v128f16        =  83,   //  128 x f16
-      v256f16        =  84,   //  256 x f16
-      v512f16        =  85,   //  256 x f16
-
-      v2bf16         =  86,   //    2 x bf16
-      v3bf16         =  87,   //    3 x bf16
-      v4bf16         =  88,   //    4 x bf16
-      v8bf16         =  89,   //    8 x bf16
-      v16bf16        =  90,   //   16 x bf16
-      v32bf16        =  91,   //   32 x bf16
-      v64bf16        =  92,   //   64 x bf16
-      v128bf16       =  93,   //  128 x bf16
-
-      v1f32          =  94,   //    1 x f32
-      v2f32          =  95,   //    2 x f32
-      v3f32          =  96,   //    3 x f32
-      v4f32          =  97,   //    4 x f32
-      v5f32          =  98,   //    5 x f32
-      v6f32          =  99,   //    6 x f32
-      v7f32          = 100,   //    7 x f32
-      v8f32          = 101,   //    8 x f32
-      v16f32         = 102,   //   16 x f32
-      v32f32         = 103,   //   32 x f32
-      v64f32         = 104,   //   64 x f32
-      v128f32        = 105,   //  128 x f32
-      v256f32        = 106,   //  256 x f32
-      v512f32        = 107,   //  512 x f32
-      v1024f32       = 108,   // 1024 x f32
-      v2048f32       = 109,   // 2048 x f32
-
-      v1f64          = 110,   //    1 x f64
-      v2f64          = 111,   //    2 x f64
-      v3f64          = 112,   //    3 x f64
-      v4f64          = 113,   //    4 x f64
-      v8f64          = 114,   //    8 x f64
-      v16f64         = 115,   //   16 x f64
-      v32f64         = 116,   //   32 x f64
-      v64f64         = 117,   //   64 x f64
-      v128f64        = 118,   //  128 x f64
-      v256f64        = 119,   //  256 x f64
+      v1f16          =  77,   //    1 x f16
+      v2f16          =  78,   //    2 x f16
+      v3f16          =  79,   //    3 x f16
+      v4f16          =  80,   //    4 x f16
+      v8f16          =  81,   //    8 x f16
+      v16f16         =  82,   //   16 x f16
+      v32f16         =  83,   //   32 x f16
+      v64f16         =  84,   //   64 x f16
+      v128f16        =  85,   //  128 x f16
+      v256f16        =  86,   //  256 x f16
+      v512f16        =  87,   //  256 x f16
+
+      v2bf16         =  88,   //    2 x bf16
+      v3bf16         =  89,   //    3 x bf16
+      v4bf16         =  90,   //    4 x bf16
+      v8bf16         =  91,   //    8 x bf16
+      v16bf16        =  92,   //   16 x bf16
+      v32bf16        =  93,   //   32 x bf16
+      v64bf16        =  94,   //   64 x bf16
+      v128bf16       =  95,   //  128 x bf16
+
+      v1f32          =  96,   //    1 x f32
+      v2f32          =  97,   //    2 x f32
+      v3f32          =  98,   //    3 x f32
+      v4f32          =  99,   //    4 x f32
+      v5f32          = 100,   //    5 x f32
+      v6f32          = 101,   //    6 x f32
+      v7f32          = 102,   //    7 x f32
+      v8f32          = 103,   //    8 x f32
+      v16f32         = 104,   //   16 x f32
+      v32f32         = 105,   //   32 x f32
+      v64f32         = 106,   //   64 x f32
+      v128f32        = 107,   //  128 x f32
+      v256f32        = 108,   //  256 x f32
+      v512f32        = 109,   //  512 x f32
+      v1024f32       = 110,   // 1024 x f32
+      v2048f32       = 111,   // 2048 x f32
+
+      v1f64          = 112,   //    1 x f64
+      v2f64          = 113,   //    2 x f64
+      v3f64          = 114,   //    3 x f64
+      v4f64          = 115,   //    4 x f64
+      v8f64          = 116,   //    8 x f64
+      v16f64         = 117,   //   16 x f64
+      v32f64         = 118,   //   32 x f64
+      v64f64         = 119,   //   64 x f64
+      v128f64        = 120,   //  128 x f64
+      v256f64        = 121,   //  256 x f64
 
       FIRST_FP_FIXEDLEN_VECTOR_VALUETYPE = v1f16,
       LAST_FP_FIXEDLEN_VECTOR_VALUETYPE = v256f64,
@@ -185,68 +189,68 @@
       FIRST_FIXEDLEN_VECTOR_VALUETYPE = v1i1,
       LAST_FIXEDLEN_VECTOR_VALUETYPE = v256f64,
 
-      nxv1i1         = 120,   // n x  1 x i1
-      nxv2i1         = 121,   // n x  2 x i1
-      nxv4i1         = 122,   // n x  4 x i1
-      nxv8i1         = 123,   // n x  8 x i1
-      nxv16i1        = 124,   // n x 16 x i1
-      nxv32i1        = 125,   // n x 32 x i1
-      nxv64i1        = 126,   // n x 64 x i1
-
-      nxv1i8         = 127,   // n x  1 x i8
-      nxv2i8         = 128,   // n x  2 x i8
-      nxv4i8         = 129,   // n x  4 x i8
-      nxv8i8         = 130,   // n x  8 x i8
-      nxv16i8        = 131,   // n x 16 x i8
-      nxv32i8        = 132,   // n x 32 x i8
-      nxv64i8        = 133,   // n x 64 x i8
-
-      nxv1i16        = 134,  // n x  1 x i16
-      nxv2i16        = 135,  // n x  2 x i16
-      nxv4i16        = 136,  // n x  4 x i16
-      nxv8i16        = 137,  // n x  8 x i16
-      nxv16i16       = 138,  // n x 16 x i16
-      nxv32i16       = 139,  // n x 32 x i16
-
-      nxv1i32        = 140,  // n x  1 x i32
-      nxv2i32        = 141,  // n x  2 x i32
-      nxv4i32        = 142,  // n x  4 x i32
-      nxv8i32        = 143,  // n x  8 x i32
-      nxv16i32       = 144,  // n x 16 x i32
-      nxv32i32       = 145,  // n x 32 x i32
-
-      nxv1i64        = 146,  // n x  1 x i64
-      nxv2i64        = 147,  // n x  2 x i64
-      nxv4i64        = 148,  // n x  4 x i64
-      nxv8i64        = 149,  // n x  8 x i64
-      nxv16i64       = 150,  // n x 16 x i64
-      nxv32i64       = 151,  // n x 32 x i64
+      nxv1i1         = 122,   // n x  1 x i1
+      nxv2i1         = 123,   // n x  2 x i1
+      nxv4i1         = 124,   // n x  4 x i1
+      nxv8i1         = 125,   // n x  8 x i1
+      nxv16i1        = 126,   // n x 16 x i1
+      nxv32i1        = 127,   // n x 32 x i1
+      nxv64i1        = 128,   // n x 64 x i1
+
+      nxv1i8         = 129,   // n x  1 x i8
+      nxv2i8         = 130,   // n x  2 x i8
+      nxv4i8         = 131,   // n x  4 x i8
+      nxv8i8         = 132,   // n x  8 x i8
+      nxv16i8        = 133,   // n x 16 x i8
+      nxv32i8        = 134,   // n x 32 x i8
+      nxv64i8        = 135,   // n x 64 x i8
+
+      nxv1i16        = 136,  // n x  1 x i16
+      nxv2i16        = 137,  // n x  2 x i16
+      nxv4i16        = 138,  // n x  4 x i16
+      nxv8i16        = 139,  // n x  8 x i16
+      nxv16i16       = 140,  // n x 16 x i16
+      nxv32i16       = 141,  // n x 32 x i16
+
+      nxv1i32        = 142,  // n x  1 x i32
+      nxv2i32        = 143,  // n x  2 x i32
+      nxv4i32        = 144,  // n x  4 x i32
+      nxv8i32        = 145,  // n x  8 x i32
+      nxv16i32       = 146,  // n x 16 x i32
+      nxv32i32       = 147,  // n x 32 x i32
+
+      nxv1i64        = 148,  // n x  1 x i64
+      nxv2i64        = 149,  // n x  2 x i64
+      nxv4i64        = 150,  // n x  4 x i64
+      nxv8i64        = 151,  // n x  8 x i64
+      nxv16i64       = 152,  // n x 16 x i64
+      nxv32i64       = 153,  // n x 32 x i64
 
       FIRST_INTEGER_SCALABLE_VECTOR_VALUETYPE = nxv1i1,
       LAST_INTEGER_SCALABLE_VECTOR_VALUETYPE = nxv32i64,
 
-      nxv1f16        = 152,  // n x  1 x f16
-      nxv2f16        = 153,  // n x  2 x f16
-      nxv4f16        = 154,  // n x  4 x f16
-      nxv8f16        = 155,  // n x  8 x f16
-      nxv16f16       = 156,  // n x 16 x f16
-      nxv32f16       = 157,  // n x 32 x f16
-
-      nxv1bf16       = 158,  // n x  1 x bf16
-      nxv2bf16       = 159,  // n x  2 x bf16
-      nxv4bf16       = 160,  // n x  4 x bf16
-      nxv8bf16       = 161,  // n x  8 x bf16
-
-      nxv1f32        = 162,  // n x  1 x f32
-      nxv2f32        = 163,  // n x  2 x f32
-      nxv4f32        = 164,  // n x  4 x f32
-      nxv8f32        = 165,  // n x  8 x f32
-      nxv16f32       = 166,  // n x 16 x f32
-
-      nxv1f64        = 167,  // n x  1 x f64
-      nxv2f64        = 168,  // n x  2 x f64
-      nxv4f64        = 169,  // n x  4 x f64
-      nxv8f64        = 170,  // n x  8 x f64
+      nxv1f16        = 154,  // n x  1 x f16
+      nxv2f16        = 155,  // n x  2 x f16
+      nxv4f16        = 156,  // n x  4 x f16
+      nxv8f16        = 157,  // n x  8 x f16
+      nxv16f16       = 158,  // n x 16 x f16
+      nxv32f16       = 159,  // n x 32 x f16
+
+      nxv1bf16       = 160,  // n x  1 x bf16
+      nxv2bf16       = 161,  // n x  2 x bf16
+      nxv4bf16       = 162,  // n x  4 x bf16
+      nxv8bf16       = 163,  // n x  8 x bf16
+
+      nxv1f32        = 164,  // n x  1 x f32
+      nxv2f32        = 165,  // n x  2 x f32
+      nxv4f32        = 166,  // n x  4 x f32
+      nxv8f32        = 167,  // n x  8 x f32
+      nxv16f32       = 168,  // n x 16 x f32
+
+      nxv1f64        = 169,  // n x  1 x f64
+      nxv2f64        = 170,  // n x  2 x f64
+      nxv4f64        = 171,  // n x  4 x f64
+      nxv8f64        = 172,  // n x  8 x f64
 
       FIRST_FP_SCALABLE_VECTOR_VALUETYPE = nxv1f16,
       LAST_FP_SCALABLE_VECTOR_VALUETYPE = nxv8f64,
@@ -257,23 +261,26 @@
       FIRST_VECTOR_VALUETYPE = v1i1,
       LAST_VECTOR_VALUETYPE  = nxv8f64,
 
-      x86mmx         = 171,    // This is an X86 MMX value
+      x86mmx         = 173,    // This is an X86 MMX value
 
-      Glue           = 172,    // This glues nodes together during pre-RA sched
+      Glue           = 174,    // This glues nodes together during pre-RA sched
 
-      isVoid         = 173,    // This has no value
+      isVoid         = 175,    // This has no value
 
-      Untyped        = 174,    // This value takes a register, but has
+      Untyped        = 176,    // This value takes a register, but has
                                // unspecified type.  The register class
                                // will be determined by the opcode.
 
-      funcref        = 175,    // WebAssembly's funcref type
-      externref      = 176,    // WebAssembly's externref type
-      x86amx         = 177,    // This is an X86 AMX value
-      i64x8          = 178,    // 8 Consecutive GPRs (AArch64)
+      funcref        = 177,    // WebAssembly's funcref type
+      externref      = 178,    // WebAssembly's externref type
+      x86amx         = 179,    // This is an X86 AMX value
+      i64x8          = 180,    // 8 Consecutive GPRs (AArch64)
+
+      i2             = 181,    // This is a 2 bit integer value
+      i4             = 182,    // This is a 4 bit integer value
 
       FIRST_VALUETYPE =  1,    // This is always the beginning of the list.
-      LAST_VALUETYPE = i64x8,  // This always remains at the end of the list.
+      LAST_VALUETYPE = i4,     // This always remains at the end of the list.
       VALUETYPE_SIZE = LAST_VALUETYPE + 1,
 
       // This is the current maximum for LAST_VALUETYPE.
@@ -415,10 +422,11 @@
     /// Return true if this is a 256-bit vector type.
     bool is256BitVector() const {
       return (SimpleTy == MVT::v16f16 || SimpleTy == MVT::v16bf16 ||
-              SimpleTy == MVT::v8f32  || SimpleTy == MVT::v4f64   ||
-              SimpleTy == MVT::v32i8  || SimpleTy == MVT::v16i16  ||
-              SimpleTy == MVT::v8i32  || SimpleTy == MVT::v4i64   ||
-              SimpleTy == MVT::v256i1);
+              SimpleTy == MVT::v8f32 || SimpleTy == MVT::v4f64 ||
+              SimpleTy == MVT::v32i8 || SimpleTy == MVT::v16i16 ||
+              SimpleTy == MVT::v8i32 || SimpleTy == MVT::v4i64 ||
+              SimpleTy == MVT::v256i1 || SimpleTy == MVT::v128i2 ||
+              SimpleTy == MVT::v64i4);
     }
 
     /// Return true if this is a 512-bit vector type.
@@ -538,6 +546,10 @@
       case nxv16i1:
       case nxv32i1:
       case nxv64i1: return i1;
+      case v128i2:
+        return i2;
+      case v64i4:
+        return i4;
       case v1i8:
       case v2i8:
       case v4i8:
@@ -705,6 +717,7 @@
       case v256f32:
       case v256f64: return 256;
       case v128i1:
+      case v128i2:
       case v128i8:
       case v128i16:
       case v128i32:
@@ -714,6 +727,7 @@
       case v128f32:
       case v128f64: return 128;
       case v64i1:
+      case v64i4:
       case v64i8:
       case v64i16:
       case v64i32:
@@ -883,8 +897,10 @@
       case i1:
       case v1i1: return TypeSize::Fixed(1);
       case nxv1i1: return TypeSize::Scalable(1);
+      case i2:
       case v2i1: return TypeSize::Fixed(2);
       case nxv2i1: return TypeSize::Scalable(2);
+      case i4:
       case v4i1: return TypeSize::Fixed(4);
       case nxv4i1: return TypeSize::Scalable(4);
       case i8  :
@@ -977,6 +993,8 @@
       case v7i32:
       case v7f32: return TypeSize::Fixed(224);
       case v256i1:
+      case v128i2:
+      case v64i4:
       case v32i8:
       case v16i16:
       case v8i32:
@@ -1171,6 +1189,10 @@
         return (MVT::SimpleValueType)(MVT::INVALID_SIMPLE_VALUE_TYPE);
       case 1:
         return MVT::i1;
+      case 2:
+        return MVT::i2;
+      case 4:
+        return MVT::i4;
       case 8:
         return MVT::i8;
       case 16:
@@ -1201,6 +1223,14 @@
         if (NumElements == 512)  return MVT::v512i1;
         if (NumElements == 1024) return MVT::v1024i1;
         break;
+      case MVT::i2:
+        if (NumElements == 128)
+          return MVT::v128i2;
+        break;
+      case MVT::i4:
+        if (NumElements == 64)
+          return MVT::v64i4;
+        break;
       case MVT::i8:
         if (NumElements == 1)   return MVT::v1i8;
         if (NumElements == 2)   return MVT::v2i8;
Index: llvm/lib/CodeGen/ValueTypes.cpp
===================================================================
--- llvm/lib/CodeGen/ValueTypes.cpp
+++ llvm/lib/CodeGen/ValueTypes.cpp
@@ -229,6 +229,10 @@
     return FixedVectorType::get(Type::getInt1Ty(Context), 512);
   case MVT::v1024i1:
     return FixedVectorType::get(Type::getInt1Ty(Context), 1024);
+  case MVT::v128i2:
+    return FixedVectorType::get(Type::getIntNTy(Context, 2), 128);
+  case MVT::v64i4:
+    return FixedVectorType::get(Type::getIntNTy(Context, 4), 64);
   case MVT::v1i8:
     return FixedVectorType::get(Type::getInt8Ty(Context), 1);
   case MVT::v2i8:
Index: llvm/lib/IR/Function.cpp
===================================================================
--- llvm/lib/IR/Function.cpp
+++ llvm/lib/IR/Function.cpp
@@ -927,25 +927,25 @@
 enum IIT_Info {
   // Common values should be encoded with 0-15.
   IIT_Done = 0,
-  IIT_I1   = 1,
-  IIT_I8   = 2,
-  IIT_I16  = 3,
-  IIT_I32  = 4,
-  IIT_I64  = 5,
-  IIT_F16  = 6,
-  IIT_F32  = 7,
-  IIT_F64  = 8,
-  IIT_V2   = 9,
-  IIT_V4   = 10,
-  IIT_V8   = 11,
-  IIT_V16  = 12,
-  IIT_V32  = 13,
-  IIT_PTR  = 14,
-  IIT_ARG  = 15,
+  IIT_I1 = 1,
+  IIT_I8 = 2,
+  IIT_I16 = 3,
+  IIT_I32 = 4,
+  IIT_I64 = 5,
+  IIT_F16 = 6,
+  IIT_F32 = 7,
+  IIT_F64 = 8,
+  IIT_V2 = 9,
+  IIT_V4 = 10,
+  IIT_V8 = 11,
+  IIT_V16 = 12,
+  IIT_V32 = 13,
+  IIT_PTR = 14,
+  IIT_ARG = 15,
 
   // Values from 16+ are only encodable with the inefficient encoding.
-  IIT_V64  = 16,
-  IIT_MMX  = 17,
+  IIT_V64 = 16,
+  IIT_MMX = 17,
   IIT_TOKEN = 18,
   IIT_METADATA = 19,
   IIT_EMPTYSTRUCT = 20,
@@ -956,7 +956,7 @@
   IIT_EXTEND_ARG = 25,
   IIT_TRUNC_ARG = 26,
   IIT_ANYPTR = 27,
-  IIT_V1   = 28,
+  IIT_V1 = 28,
   IIT_VARARG = 29,
   IIT_HALF_VEC_ARG = 30,
   IIT_SAME_VEC_WIDTH_ARG = 31,
@@ -979,12 +979,14 @@
   IIT_BF16 = 48,
   IIT_STRUCT9 = 49,
   IIT_V256 = 50,
-  IIT_AMX  = 51,
+  IIT_AMX = 51,
   IIT_PPCF128 = 52,
   IIT_V3 = 53,
   IIT_EXTERNREF = 54,
   IIT_FUNCREF = 55,
   IIT_ANYPTR_TO_ELT = 56,
+  IIT_I2 = 57,
+  IIT_I4 = 58,
 };
 
 static void DecodeIITType(unsigned &NextElt, ArrayRef<unsigned char> Infos,
@@ -1037,6 +1039,12 @@
   case IIT_I1:
     OutputTable.push_back(IITDescriptor::get(IITDescriptor::Integer, 1));
     return;
+  case IIT_I2:
+    OutputTable.push_back(IITDescriptor::get(IITDescriptor::Integer, 2));
+    return;
+  case IIT_I4:
+    OutputTable.push_back(IITDescriptor::get(IITDescriptor::Integer, 4));
+    return;
   case IIT_I8:
     OutputTable.push_back(IITDescriptor::get(IITDescriptor::Integer, 8));
     return;
Index: llvm/utils/TableGen/CodeGenTarget.cpp
===================================================================
--- llvm/utils/TableGen/CodeGenTarget.cpp
+++ llvm/utils/TableGen/CodeGenTarget.cpp
@@ -56,6 +56,10 @@
   switch (T) {
   case MVT::Other:    return "MVT::Other";
   case MVT::i1:       return "MVT::i1";
+  case MVT::i2:
+    return "MVT::i2";
+  case MVT::i4:
+    return "MVT::i4";
   case MVT::i8:       return "MVT::i8";
   case MVT::i16:      return "MVT::i16";
   case MVT::i32:      return "MVT::i32";
@@ -88,6 +92,10 @@
   case MVT::v256i1:   return "MVT::v256i1";
   case MVT::v512i1:   return "MVT::v512i1";
   case MVT::v1024i1:  return "MVT::v1024i1";
+  case MVT::v128i2:
+    return "MVT::v128i2";
+  case MVT::v64i4:
+    return "MVT::v64i4";
   case MVT::v1i8:     return "MVT::v1i8";
   case MVT::v2i8:     return "MVT::v2i8";
   case MVT::v4i8:     return "MVT::v4i8";
Index: llvm/utils/TableGen/IntrinsicEmitter.cpp
===================================================================
--- llvm/utils/TableGen/IntrinsicEmitter.cpp
+++ llvm/utils/TableGen/IntrinsicEmitter.cpp
@@ -196,25 +196,25 @@
 enum IIT_Info {
   // Common values should be encoded with 0-15.
   IIT_Done = 0,
-  IIT_I1   = 1,
-  IIT_I8   = 2,
-  IIT_I16  = 3,
-  IIT_I32  = 4,
-  IIT_I64  = 5,
-  IIT_F16  = 6,
-  IIT_F32  = 7,
-  IIT_F64  = 8,
-  IIT_V2   = 9,
-  IIT_V4   = 10,
-  IIT_V8   = 11,
-  IIT_V16  = 12,
-  IIT_V32  = 13,
-  IIT_PTR  = 14,
-  IIT_ARG  = 15,
+  IIT_I1 = 1,
+  IIT_I8 = 2,
+  IIT_I16 = 3,
+  IIT_I32 = 4,
+  IIT_I64 = 5,
+  IIT_F16 = 6,
+  IIT_F32 = 7,
+  IIT_F64 = 8,
+  IIT_V2 = 9,
+  IIT_V4 = 10,
+  IIT_V8 = 11,
+  IIT_V16 = 12,
+  IIT_V32 = 13,
+  IIT_PTR = 14,
+  IIT_ARG = 15,
 
   // Values from 16+ are only encodable with the inefficient encoding.
-  IIT_V64  = 16,
-  IIT_MMX  = 17,
+  IIT_V64 = 16,
+  IIT_MMX = 17,
   IIT_TOKEN = 18,
   IIT_METADATA = 19,
   IIT_EMPTYSTRUCT = 20,
@@ -225,7 +225,7 @@
   IIT_EXTEND_ARG = 25,
   IIT_TRUNC_ARG = 26,
   IIT_ANYPTR = 27,
-  IIT_V1   = 28,
+  IIT_V1 = 28,
   IIT_VARARG = 29,
   IIT_HALF_VEC_ARG = 30,
   IIT_SAME_VEC_WIDTH_ARG = 31,
@@ -248,12 +248,14 @@
   IIT_BF16 = 48,
   IIT_STRUCT9 = 49,
   IIT_V256 = 50,
-  IIT_AMX  = 51,
+  IIT_AMX = 51,
   IIT_PPCF128 = 52,
   IIT_V3 = 53,
   IIT_EXTERNREF = 54,
   IIT_FUNCREF = 55,
   IIT_ANYPTR_TO_ELT = 56,
+  IIT_I2 = 57,
+  IIT_I4 = 58,
 };
 
 static void EncodeFixedValueType(MVT::SimpleValueType VT,
@@ -291,6 +293,10 @@
     return Sig.push_back(IIT_EXTERNREF);
   case MVT::funcref:
     return Sig.push_back(IIT_FUNCREF);
+  case MVT::i2:
+    return Sig.push_back(IIT_I2);
+  case MVT::i4:
+    return Sig.push_back(IIT_I4);
   }
 }